Lessons From the Titans

From the productivity of the factories to every contract term, they emphasized cash flow. And that cash flow allowed for five large bets on future growth: (1) air travel, (2) gas-powered electricity generation, (3) global healthcare expansion, (4) television advertising, and (5) financial services.

Money seemed to pass back and forth within GE businesses, for no clear reason special-purpose entities were created, and loans were originated and then syndicated, only to fall off the balance sheet, all while creating some sort of paper profit gain. Complex derivative products were used. Long-term debt was swapped for short-term, fixed-rate debt for floating and vice versa. Currencies were bought and sold, all in the name of “hedging,” but there seemed to be no consistency if that was the case. Special leasing arrangements were sometimes used so that even shipping a unit across the street would constitute a sale and revenue booked. Tax gains and losses were created when needed to smooth results. GE would fight the IRS on every penny. Most quarters saw some sort of “tax gain,” and that gain was buried in a financial statement category called “other.” It was beyond complex. It was financial engineering at its most extreme—perhaps on a scale never seen before and not since. And there was no one outside the walls of corporate HQ that had any ability to follow all the pieces. Those of us who followed the company for a living were just kidding ourselves that we understood all the financials.

But little did investors know that GE retained substantial tail-liability risk in several of its deals. Long-term-care insurance had the biggest impact. As losses on those policies piled up, despite the company’s best efforts to hide them, GE eventually took a massive $15 billion hit to reserves in 2018—a size that few could have ever imagined possible, particularly given the fact that many investors were led to believe GE had completely exited the insurance business via the IPO of Genworth in 2004 and asset sales to Swiss Re in 2005. (See

The context is astounding. NBCU, which equated to less than 20 percent of GE’s earnings in most years prior to its sale, is now worth about all of what remains at GE combined. GE sold at the absolute bottom, leaving billions on the table. When I later brought up that conversation to Immelt in front of a small group of investors at an annual industry conference, he asked me, “When did you become such an asshole?”

These assets were a bit more expensive than the traditional cyclical deal book, but Culp took on low-cost debt as an offset—not just from lower interest rates, but from the rating agencies’ realization that Danaher was a cash flow machine. Another benefit from reducing cyclicality was that the market was willing to pay a premium for a Danaher stock with less variability. Culp saw that trend early—and got the Rales brothers and the board to agree to a strategy of paying more for the right kind of assets.

Most corporate buyers fixate on either higher-growth game-changing transactions or deals that serve to consolidate their markets. But Culp went after solid assets—healthcare, dental, and testing equipment firms that often flew a bit under the radar—not high growth, and often very niche, which meant they didn’t fit well with other strategic buyers.

In a world where capital was becoming easier to access, private equity was increasingly driving up deal valuations, which meant that execution levels needed to be higher than ever. The corporate folks involved in deal execution and integration needed to be exceptional, so Culp sought to build up that talent level to match the increasing needs of the organization.

Dan Comas came to Danaher in 1991 in an M&A role and became CFO in 2005. He was well pedigreed with degrees, from Georgetown and Stanford, and had a reputation for intelligence and a rational thought process. Throughout his career at Danaher he closed more than 250 deals worth more than $45 billion, an astounding level of volume. His contributions to M&A alone could be considered his greatest achievement.

“Now that the deal is closed, please tell us everything that you failed to tell us during diligence.” He believed that things always go wrong in a deal; what distinguishes the good from the bad deals are the speed and permanence of the fixes.

he could deliver and present the reliable financial numbers that investors favored: straightforward results with limited surprises quarter to quarter. Beyond that, Anderson helped Cote to execute on a mile-long to-do list internally. There was no shortage of work.

He took the cash freed up by HOS and its Lean foundation, used it to pay down debt, and got back to minimum funding levels for the pension plan, all while reaching quick settlements on much of the asbestos and environmental damages. Wall Street eventually noticed the sharply falling risk profile.

Most CFOs back then focused on boosting cash flow through higher earnings, but Anderson saw multipliers on that opportunity. He was astounded by the cash tied up in working capital: billions of dollars in inventory, accounts receivable, and other areas.

Prior to the Cote/Anderson era, the company converted less than 80 percent of its net income to cash flow, was starving its operations of investment, and was underfunding its pension.

They believed the company’s stock was deeply undervalued, so they issued new shares and put them in the underfunded pension plan. Investors hated the move because it diluted their ownership stake, but it went a long way to solving the immediate liability crisis.

Principal among these concerns was underinvestment in new products and technologies, as R&D projects were being shelved in favor of maintaining or expanding shorter-term profits. In David’s first five years as CEO, company spending on R&D totaled ~5 percent of revenue on average. However, this spending fell to roughly 3 percent by the time he retired.

The pursuit of cost minimization and profit maximization is great in principle. Efficiency isn’t a bad thing, but taken too far it comes with consequences. Lower costs applied to well-understood, preexisting businesses are pretty straightforward, as they fall straight to the bottom line. However, if new business opportunities are judged versus legacy margin rates, investments in projects that offer growth potential, attractive returns on capital, or even more loosely defined “strategic value” may never materialize.

The IAE transaction was structured in a way that pulled Rolls-Royce’s share of earnings onto UTC’s financial statements, while deferring royalty payments into future years. This boosted short-term EPS at the expense of future cash flow.

However, the company’s cash flows grew approximately half as fast (see Figures 6.6 and 6.7), suggesting that the reported earnings were being overstated. This proved to be the case, as some benefits that had been counted as earnings didn’t have cash flow to go along with them.

CAT makes decent money selling machines, but selling repair parts for those machines for the next 15 to 20 years is more profitable. Giving up market share hurt long-term as well as near-term profits, and that drove Oberhelman nuts.

That had a terrible consequence. It led to double ordering: a dealership might want two machines to sell to its customers, but historically it had only gotten one in a timely manner. So now the dealer orders four, figuring it will get two and will cancel the other orders later if equipment isn’t moving.

Critical to Jellison’s contrarian view was the mathematical reality that stock market P/E ratios directly correlate to returns on the underlying asset base of the entity, meaning that P/E ratios usually rise as the returns on the underlying assets grow. The very highest P/E multiples are at companies whose business models have few physical assets to maintain and high margins on that low asset base. Companies with low asset intensity and high margins typically generate lots of cash. And companies with these characteristics that also grow are particularly attractive, as each incremental dollar of growth drives returns even higher. This is a concept Jellison called CRI—cash return on investment—and its correlation with valuation was central to his vision. The best example of a high-CRI sector would be software. The best example of a low-CRI sector would be automotive.

With the high volume of potential M&A targets in the $100 million–$500 million range, Jellison could afford to be picky with his deals. He had three requirements: (1) lower asset intensity than ROP’s existing portfolio, (2) good businesses within niche industries, and (3) excellent management. He would walk away from any deal that lacked all three.

The valuation for an asset-heavy company with potential tail liabilities was not much less than the valuation for a software company.

Jellison could take that 5 percent unit growth, and through operational improvement, he could translate it into nearly 10 percent profit growth. He then could utilize cash to purchase similar assets that would add another 5 to 10 percent to the top line, translating over time to 15 to 20 percent annualized profit growth. He did this without using equity and maintaining investment-grade debt.

If we learn nothing else from the Jellison experience, it’s that valuation mispricings can persist over a very long period of time. Even later in Jellison’s career, he found reasonably priced software assets at valuations not much more than those paid for more asset-intensive businesses. We have seen those spreads widen over the last few years, in some cases exponentially. But for north of 15 years, Jellison was able to capitalize on a marketwide opportunity to invest in niche businesses that few would take the time to consider.

He loved the job, and he loved the game overall, which he viewed as a competition. He wanted to beat the big guys—the GEs, the 3Ms, and the Emersons of the world. An outcast from that world, he was considered not polished enough, with opinions that were too strong. He had a chip on his shoulder. He wanted to win by having a better operating framework, proven by his sector-high margins. He wanted to invest cash at high returns and drive the stock price further up. He benchmarked his company against others, almost to a level of infatuation.

The Roper process seems to cut down the risk profile to an almost uncanny extent. Of the 50+ deals we observed Jellison do in his 17 years, we can name only a couple that didn’t work out, and those were just modest disappointments.

The risk profile was also lowered, because there was no such thing as “integration risk.” Roper assets are autonomous, never integrated. There are never social or cultural issues, because the existing entity is maintained. There isn’t a mass exodus risk, because managers are locked down before the deal closes. This could be a bigger part of the deal success story than we give it credit.

Because all his acquisitions were high gross margin and low capital intensity, paying on profit growth meant that managers really had to get only the top-line growth measure right. That was the most powerful lever they could pull. The incentive to grow with and above the market was already there.

In an effort to knock Howley off balance, the investor forcefully asked him why he should believe in the company following Howley’s recent sale of millions of dollars of stock. Howley stunned me when he sat up in his chair, put his elbows on the boardroom table, looked the guy in the eye, and said: “This may come as a surprise to you, but I’m in this for the money. I haven’t had many chances to sell stock under private equity ownership, and now I do. My wife wants a beach house. So we’re gonna get a beach house. You can believe what you want, but I’m not going anywhere. I’ve got more money to make, and if you choose to, you can make it with me.” This told me almost everything that I needed to know about TransDigm.

These companies build on prior engineering and manufacturing know-how, existing capital and certification investment, and track records of safety and reliability to create mini-monopolies that enjoy sales of specific aircraft parts for decades with low levels of competition and reinvestment. These competitive and regulatory moats make the aviation supply chain home to a lot of good businesses, the best of which make money selling low-cost, relatively high-priced spare parts for the ~30 years that an aircraft is in service.

The older an airplane gets, the more expensive it becomes to operate. This leads to shrinking fleets, as older, more expensive planes are retired or used less frequently in favor of newer, more efficient aircraft. By this time in an aircraft’s life cycle, the likelihood of an alternative part being certified is increasingly low—new entrants struggle to justify the engineering and regulatory investment required to sell a new part into a shrinking market against an already established and trusted competitor. This allows the legacy manufacturers to push through annual price increases more aggressively. This pricing power more than offsets the declines in volume and justifies keeping production lines running efficiently, as incoming order velocity slows. These parts generate a ton of profit.

At technology companies, growth is often emphasized above anything else. However, abandoning high-margin and high-return products when they slow or stop growing leaves significant value on the table.

Venture capital (VC) funding for industrial markets, under a billion dollars 10 years ago, exceeded $40 billion in 2018. Total R&D spent by all public industrial companies is now around $90 billion, so in a few short years VC investment has reached close to half that level. (See Figure 10.7.) Most of this influx of funding aims at disruptive innovations, while most standard R&D is incremental. Accordingly, the tech world is now outspending the industrial world on real disruption by a wide margin and with greater efficiency.

The new metrics were simple: “return on controllable assets” with a hurdle rate to reflect that assets cost money and need to generate a minimum return before any bonuses could be paid.

GE gas turbines now generate a third of all the electricity in the world.

In gobbling up a key competitor, GE was forced to create a new one, which crushed all the benefits of consolidation in the first place. But GE pressed on, even raising its bid to $17 billion ($11 billion of it in cash) in order to beat out a joint bid from Siemens and Mitsubishi.

We just didn’t know what many GE insiders knew: Alstom was nearly bankrupt.

In fact, most of our successful case studies had notable talent in the CFO seat. And just the opposite in our failure case studies.

The CFO job is underappreciated by a country mile in almost every great corporate story.

We find CFOs heavily influencing strategic direction, offering eyes and ears on disruption risks and competitive context that operating executives often miss.

Anderson’s strong stance on earnings quality wasn’t popular, and he had to replace most of his 70+ reports.

Pay top talent not just for the job they do today, but for the job they will be offered tomorrow.

Forecasting feels like a science, but it isn’t.

The answer came quickly. The growth did materialize, and to this day, water companies continue to change out old meters that require in-person reading with wireless ones that are run off radio frequency identification (RFID).

Roper wants all the pros and cons out on the table, well before it goes down the bidding path. Easier said than done, but Jellison had a way of finding problems up front and walking away from the transaction. This is perhaps why his hit rate was so high. There was extreme process in his diligence.

Margin expansion is great, but it only really counts if it is systematic and repeatable.

Like jet engines, power turbines have low margins up front, but they last for decades with years of high-margin replacement parts and service revenues.

GE’s acquisition of Baker Hughes a year later seemed to provide another tailwind, all assuming that the demand for oil and gas, along with jet engines, healthcare, and, of course, gas turbines, would stay strong.

However, understanding the technological specifications and delivering manufacturing excellence are very different things.

Fewer still are the industrial companies that have successfully adapted to the rise of digital technology in the industrial sector. The systems of rapid innovation in the technology world are just too different, as are the networks of engineers, startups, and acquisitions required to keep pace.

The airplane fiasco reported by the Wall Street Journal in 2017, where a backup jet followed him on business trips in case the main jet had problems, exposed just a small piece of the excess. In selling NBC in 2009, for example, he had an entire golf course shut down so he and Brian Roberts of Comcast could negotiate in private while playing.

But as he reduced the company’s exposure to financial services, he needed something to replace the lost earnings. He settled on two big initiatives: digital transformation and the acquisition of a large French power infrastructure company named Alstom.

Unlike Welch’s focused bets, Immelt’s digital strategy had no specific digital product in mind, at least not one that customers were currently looking for. Instead it was more of a concept. “GE would become the backbone to the entire industrial internet” sounded sexy, but it wasn’t clear that it was necessary or even all that useful.

A turbine repair under warranty, for example, could cost a few hundred thousand dollars while the turbine remained at GE, but cost many millions of dollars once installed at the customer site.

Electrical Products Group Conference,

Not knowing how much money the company was making or losing on each customer’s aircraft demonstrated a lack of financial rigor that he found simply unacceptable. The complexity and scale of aircraft manufacturing leave little room for error, and a program’s success or failure cannot be left to chance.

Boeing believed it was untouchable, and by consistently underestimating the potential for anyone, let alone Airbus, to challenge the company’s market position, Boeing enabled a formidable competitor’s rise.

I asked him bluntly what he thought had gone so wrong, and he candidly said the company had too often pushed technology development much further than customers were willing to compensate Boeing for.

Under those conditions, with two otherwise identical businesses, buying the highly cyclical company is going to generate less value for Danaher than the steadier, less cyclical one. Nothing was more frustrating to Culp than getting a company fully ramped up on Lean, only to see a recession unwind so much of the hard work, so he pushed for a change in the type of assets that he wanted to acquire.

Separately Culp came to understand that DBS worked better with businesses with a high gross margin, especially those with a big spread between gross and operating margins. The wider the spread, the more opportunity for Danaher to take costs out with DBS. Culp had learned through operating experience that it was easier to improve the gross margin of a high-margin company versus a low-margin one.

In 1991 the company tracked just about everything it could, including at least 50 financial metrics that it used for performance appraisals and compensation. Comas found it overwhelming and unhelpful, and through many kaizen events and iterations, he narrowed down the focus to eight core metrics. Just four were financial: organic growth, margins, cash flow, and ROIC. Another two of the metrics related to the customer experience: on-time delivery and quality measured by defects/million. And the last two involved people: internal job fill rate and retention. Danaher still uses these eight simple metrics to measure manager performance and benchmarks across the organization and within specific regions, metrics we would consider a best practice overall. (See Figure 4.7.) Figure 4.7: Danaher’s eight metrics. * Refers to product defect rates. Source: Danaher

On the other hand, as Danaher has sought higher gross margin assets, the cost opportunity has increasingly been outside of manufacturing, and that’s where the evolution of DBS has been most pronounced. Driving efficiency in sales and marketing then becomes crucial for the deal model overall—and Danaher has developed matching tools.

Design at Fortive, for example, is all about listening to the customer and creating only products that customers want, need, and are willing to pay for. The product development stage is all about milestones, and each design iteration is focused on simplicity.

Where many companies will pass on responsibility for each step to different groups, FBS tools get all folks involved early in the process. There is codified collaboration among groups, which include early product developers, procurement, sales/marketing, and even digital partners, in war rooms teeming with visuals that keep all parties engaged and focused.

Honeywell pivoted and reenergized around big themes: energy efficiency, productivity, and connectivity—more popularly known as the industrial internet of things (IIoT), with the high-margin software businesses that have accompanied the strategy.

Though the turnaround took north of seven years from start to finish, progress was steady, and by the end of year two the company was clearly on the right path. But getting to that point was painful and teaches the lesson that there are no shortcuts.

and the new senior leadership team got Honeywell out of some lousy businesses at decent prices. Gone were the worst of the commodity chemical assets, a large consumer auto business, and a subscale alarm monitoring service. At businesses the company couldn’t sell, costs were cut and customer contracts restructured. The company also bought some new businesses that proved a good fit and entered new markets with attractive characteristics—all while avoiding expensive targets where Honeywell couldn’t add much value.

Air conditioners aren’t sold direct from the factory. They get to the end customer through a two-step distribution system. The profits from the distribution part of the sales process aren’t particularly high, so pulling the distribution assets out of Carrier pushed margins higher simply by improving the mix.

Without that, a company is forced to try to manage by feel, using forecasting and intuition to guess demand.

Excellence in production systems can dampen the swings by controlling everything else other than demand: quality, on-time delivery, safety, and many other metrics that are independent of volume. Robust and disciplined business systems help ensure that.

The problem is, planning rarely works well. Not with cars, and certainly not with mining trucks. When demand is lower than expected, you’ve just built too many cars, and they are out there sitting in lots somewhere. Eventually production has to be cut, and cut deeper than the level of current demand, to get rid of all the extra unsold inventory.

Market share losses hurt equipment sales in one year, and then they hurt parts and aftermarket sales for the next 15 years that CAT’s machine would have been running.

The lesson from the Owens era is that the economist’s approach failed. Forecasting supply and demand levels didn’t work; expecting a continuation of weak demand, CAT lost market share, margin, and strategic position.

That touches on something we may fail to highlight enough in this book. In all the ways that smart people can complicate all forms of business, the best of the best just seem to focus on a few basic drivers of value. Roper is perhaps the best illustration of this dynamic.

Flying under the radar often meant margins twice what the main suppliers earned, and the overall investment needs were limited and easier to cut back on in recession periods. The factories were smaller, closer to the customer overall, and focused on final assembly. Jellison also saw the big players increasingly interested in larger, splashy M&A deals. Smaller, niche assets (sub-$500 million and even sub-$100 million) garnered more attractive prices. In addition, he saw that smaller company managers were usually more intimate with their businesses overall and respected within their organizations but had not been incentivized appropriately.

Boards wanted easy-to-explain combinations, regardless of whether the underlying business itself was good or bad.

When they did do acquisitions, they wanted a high-growth profile—certainly above 10 percent, and rarely near the 5 percent ballpark that Roper was more than happy to take. Jellison found the biggest mispricings in less sexy, slower, but still solid growers. These were usually software companies in highly niche markets.

AmSafe accounts for more than 95 percent of the global aircraft seat belt market. How good a business is airplane seat belts? After all, it’s only three pieces of metal and a spring. Well, you can ask TransDigm, which paid $750 million for AmSafe in 2012 and has generated a 20+ percent return on its investment. Aircraft seat belts are a fantastic business, especially in the right hands.

TransDigm is remarkably consistent in the application of what it calls its three value drivers: (1) value-based pricing, (2) productivity, and (3) profitable new business. It’s a basic formula that requires masterful execution to get right.

That means that as older planes are retired and production runs of spare parts become less frequent and more challenging to predict, TransDigm demands that it be compensated, not just for the direct cost of the part, but for the costs of keeping production lines “hot” with skilled labor and working machinery. The company takes reliability seriously, and it expects to be paid for delivering a high-quality, usable part quickly.

TransDigm’s pricing strategy remains sustainable because of three key factors: (1) regulatory, (2) economic, and (3) customer behavior.

Picture this: A business line manager at a large company looks at the list of products in her business unit and says, “I could grow this operation’s revenues faster if I could get rid of the stuff that’s shrinking.” She finds products that go on older aircraft no longer in production, where inventory turns slowly and innovation is scant. If she gets rid of these parts, her business’s overall growth rate goes up over time, and her inventory goes down. Depending on how she’s being paid, which most often is by sales growth, she’s a hero. There’s only one problem with this picture: these products often have the highest potential profit in the portfolio. They have the best pricing power, and they require little or no ongoing investment. The fact that managers were allowed to sell such product lines to Extant is crazier than anything TransDigm has ever done.

Consider how much easier it would be to sell to one large-scale customer, like Walmart, versus the intense sales strategy needed to get product into a vast multitude of independent convenience stores. In this way the rise of home centers like Home Depot and Lowe’s made it easier for new tool manufacturers to scale up, driving out much of the older, diverse hardware store and lumberyard base where tools had been sold. However, the home centers also accelerated the move to lower-cost sourcing from China.

These are all simple issues, but a mistake on any one of them can lead to cascading delays and angry customers. Discounting price or other services are some of the potential results. The right solution isn’t necessarily some complex IT project to drive better deliveries.

There’s nothing that destroys a culture faster than wasteful spending and celebrity behavior among the executives. Every example of corporate failure we know about included exactly that.

It’s a journey, not a destination, and most cultures look for explicit goal lines. The result is that people focus on that goal line, rather than continuous improvement itself, and once that goal is achieved, the improvement stops.

Visual management is a simple yet powerful tool.

Change before you expect it is probably a good thing. In fact, a strong culture may drive that change. And the opposite of success through reinvention may just be failure through stagnation.

There are no new paradigms. Danaher and Fortive employees aren’t expected to reinvent the wheel. They’re expected to make that wheel go a little faster and smoother every single day.

Cote favored hunting and fishing over golf. He drank diet soda and beer from a can and ate fast food, Kentucky Fried Chicken I recall being his favorite. He liked to ride his motorcycle on weekends and smoke cigars. Years later, he visited my home for an event I was hosting and introduced himself to my wife by saying, “Hi, I’m Dave. I run Honeywell.” She walked over to me and said, “I think he picked up his jeans from the floor, where he left them the night before.” He hardly looked the part of the game-changing leader of a critical American company.

My take is that Cote badly needed Anderson and vice versa. The turnaround was too difficult to accomplish without a strong finance head to complement those on the front lines in operations. Regardless of who did what when, Cote deserves the credit for hiring someone who compensated for his weaknesses.

Two Cote quotes I remember from that meeting: “I want to make money on the stock 10 years after I retire” and “I want people to say, ‘If I want a good leader, I hire a Honeywell guy’—except they can’t get our guys because Honeywell is such a great place to have a career.”

This is a key lesson we have seen in all other winning cases: culture as an output of incentives and the actions and direction of leadership, not an input.

Forecasting is bound to be wrong and should be done as little as possible. The better path is flexible, rapidly responding, Lean operations. The problem is that companies are just too tempted to shortcut the process. The hard part about business systems isn’t getting started, nor is it doing kaizens or talking about Lean; the hard part is staying committed. It’s all too easy for a management team to see its time filled with planning, forecasting, and budgeting and then to set systems aside to chase growth when given the chance.

An analysis of a half century of aircraft development programs suggests that bets on new aircraft pay off and make money only half of the time.

It cost more than $1.2 billion by the time it was “done,” well more than 10 times its intended selling price.

Small increments of improvement, done in steps, become large increments of improvement over time.

In the meantime, watch for Danaher and Fortive to operate with the mindset of “common sense vigorously applied.”

Cote was focused on success and was confident in his abilities. Just as important, though, he never wanted to go back to the life of struggling paycheck to paycheck.

For an airline, the price points across TransDigm’s portfolio of parts are relatively low. Seat belts cost a minuscule fraction of what a spare part for an engine costs, but an airline needs both in order to be cleared for takeoff. TransDigm’s history of consistent on-time delivery and quality performance makes it rare for customers to switch providers. The risk isn’t worth it. Airlines cannot afford to miss out on several hundred thousand dollars of revenue if flights are canceled because a $200 spare part is not delivered on time or to specification.

But perhaps that’s one of the key lessons of this chapter: success often requires a wild ride, full of bold decisions and pivots, but transformational change anchored in a solid system.

Cote used to say, “What you think, what you say, and what you do shouldn’t be three separate decisions.

After it exited PE ownership, the company retained two core beliefs from its experience: (1) thoughtful cash deployment and a well-designed capital structure were tools for creating additional value, and (2) employees should think, act, and be compensated like owners.

The rise and fall of United Technologies highlights that the best incentive systems are not static or myopically focused; they should exhibit the same dynamism as an organization’s ever-evolving set of opportunities and risks.

the case of United Rentals, the incentive puzzle wasn’t hard. Managers had been compensated on growing EBITDA. In other words, they were paid for revenue and profit growth but were not charged for the fleet they owned. That promoted a simple, pro-growth decision: buy more stuff and worry about renting it later.